Alcohol, the heart and the cardiovascular system: What do we know and where should we go?

The cardioprotective effect of alcohol has been established in medical epidemiology, mainly based on individual level cohort studies, as a very consistent finding (for meta-analyses see [1] -[3] ). Methodologically, most of the studies have some problems (e.g. [4] ), but the biological pathways have been clearly established, especially for ischaemic heart disease ([5] , [6] ; see also [7] ). However, a large study recently found a differentiation within the group of ischaemic heart disease outcomes ([8] ; see also [9] ), with alcohol impacting detrimentally on ischaemic heart disease other than myocardial infarction. As with ischaemic heart disease, a protective effect with similar pathways for ischaemic stroke has been established [10] . On the other hand, various other cardiovascular conditions are detrimentally impacted by alcohol, including haemorrhagic stroke [10] , hypertensive diseases [11] and atrial fibrillation [12] , again with clear biological pathways (see also [2] , [6] ). The picture becomes more complicated when we look into aggregate level studies and natural experiments. Aggregate level studies display mixed results between protective and detrimental effects of per capita consumption on ischaemic heart disease deaths [13] , sometimes even within one country [14] . The same is true for natural experiments and the larger group of cardiovascular deaths: during the Gorbachev reform, a decline in alcohol consumption as part of an effective government campaign was associated with lower rates of cardiovascular death [15] , whereas in Finland, an increase in alcohol consumption as the result of a large decrease in prices was also associated with lower rates of cardiovascular death [16] . Thus opposite changes in consumption led to the same result. Can we make sense of all of these seemingly contradicting results? We believe we can, by mainly separating different patterns of drinking [17] —regular low to medium volume drinking, regular heavy drinking and irregular heavy drinking (see also [18] )—and by differentiating the outcomes. In terms of drinking patterns, the results of individual level cohort studies mainly apply to cohorts who predominantly drink regularly and with few heavy drinking occasions. This results from the usual recruitment mechanisms for cohorts which stress the possibility of recontacting members for the following years [19] —perhaps best exemplified by the famous cohorts of nurses and health professionals. This recruitment feature means that the drinking style of typical cohorts reflects the favourable, regular low to medium volume regular drinking style of the middle class in high income countries, and increasingly also in some middle income countries, but not the style of the majority of people who consume alcohol globally, which is characterised more by heavy drinking occasions [20] . Thus, in cohorts or countries where this style prevails, we would expect a risk curve similar to that of the cohort studies: an overall beneficial effect of alcohol on ischaemic diseases. Conversely, a drinking pattern characterised by irregular heavy drinking occasions does not seem to confer any beneficial effect on ischaemic heart disease [21] , [22] . In Russia, as an extreme example for such a drinking style, we would expect a mixed or even overall detrimental effect on ischaemic diseases, as confirmed in nearly all analyses except for selected cohorts with more regular and moderate drinking habits. This effect on ischaemic disease will add to the detrimental effect on other cardiovascular diseases which seems to be independent of patterns of drinking, and leads to a clear detrimental overall effect of alcohol consumption on cardiovascular disease. As for the study of Herttua and colleagues, who found a protective effect on cardiovascular deaths using aggregate level data in Finland [16] , which has traditionally been classified as an irregular heavy drinking country (e.g. [23] ), we can only speculate that the majority of Finns at the ages of forty and above, and especially for ages above 65, where ischaemic disease plays a role, probably drink more like the middle class cohorts than like Russians. Surveys overall seem to confirm this picture of overall moderate drinking style, although there is still a high degree of irregular heavy drinking (e.g. [24] ). The question of regular heavy drinking is interesting. Few data exist to investigate the impact of regular heavy drinking because such a drinking pattern is relatively scarce in most populations and disproportionally missed in usual cohort studies. Evidence from cohort studies examining average heavy drinking in comparison with abstainers is mixed, with more than 60 g day−1 usually being the highest (and thus open-ended) category with the lowest number of participants. There are also some short-term experimental studies. Out of such experiments, six found a positive association on high-density lipoproteins [25] -[30] and one no association [31] . However, the impact of regular heavy consumption on platelets' aggregation seems less clear and may be in fact detrimental [32] . Epidemiologically, the effect of regular heavy drinking on ischaemic disease is mixed. It may be that the protective effect of regular heavy drinking occasions on lipids is counteracted by the detrimental effect on hypertension [11] . Drinking with or outside meals may be an important co-determinant here [2] . Adequate power to investigate such complex relationships requires very large cohorts depending on the distribution of such drinking patterns within a population with rather detailed measurement of alcohol exposure and exposure to other heart disease risk factors. As demonstrated here, we believe that we have some understanding of how alcohol impacts the cardiovascular system, and the impacts are indeed complex. We should continue to build on this understanding and its complexities further, rather than falling back to undifferentiated positions, such as trying to ‘disprove’ any beneficial effect. The understanding of the complex interplay of different patterns of drinking on different cardiovascular outcomes will not only improve our knowledge, but also will help in the prevention of cardiovascular disease burden, and chronic disease burden in general [33] . We would like to thank Dr Pia Mäkelä for her valuable comments on the interpretation of the Finnish results, and Dr Alison Ritter for her insights and helpful comments in shaping the final version.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.